In wireless sensor networks, data aggregation protocols are employed to prolong the network lifetime. However, performing data aggregation while preserving security is a challenging problem. This paper presents a polynomial regression based secure data aggregation protocol in which sensor nodes represent their sensed data as polynomial functions. Instead of their original data, sensor nodes secretly send coefficients of these polynomial functions to data aggregators. Data aggregation is performed based on these coefficients and the base station is able to extract a good approximation of the network data from the aggregation result. The security analysis and simulation results show that the proposed scheme is able to reduce the amount of data transmission in the network without compromising data confidentiality.